celu

celu(x: &Tensor<T>, alpha: C) -> Result<Tensor<C>, TensorError>

Compute $\large \text{max}(0, x) + \text{min}(0, \alpha \cdot (e^{x/\alpha} - 1))$ for all elements

Parameters:

x: Input values

alpha: Parameter controlling the saturation of negative values

Returns:

Tensor with type C

Examples:

use hpt::{ops::FloatUnaryOps, Tensor, error::TensorError};

fn main() -> Result<(), TensorError> {
    let a = Tensor::<f32>::new([-1.0]);
    let b = a.celu(1.0)?;
    println!("{}", b);  // prints: -0.6321206
    Ok(())
}

Backend Support

Backend	Supported
CPU	✅
Cuda	✅